Date:         Mon, 13 Mar 2000 21:05:12 GMT
Reply-To:     Steve Gregorich <gregorich@PSG.UCSF.EDU>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Steve Gregorich <gregorich@PSG.UCSF.EDU>
Organization: University of California, San Francisco
Subject:      Re: About Proc Factor
Content-Type: Text/Plain; charset=US-ASCII

Have you requested method=ML?

If so the model-reproduced covariance
matrix (or corrrelation matrix) will
reproduce the diagonal elements *exactly*.
This means that all non-zero residuals
will be in the off-diagonal elements.
Thus, the root-mean-square off diagonal
residuals is the same thing as the
root-mean-square residual, commonly
labeled RMR.

HTH



Steve Gregorich

>Dear Sas-users:
>
>I have a question about Proc Factor. One of the output when I performed
>this procedure is "Root Mean Square Off-diagonal Residuals".The output
>includes the over-all score and the score for each variable in the
>model. I can't find any explanation from SAS manuals about this part...
>Maybe my manuals had gotten too out of date... I only know that this
>could be one of the indexes to choose the number of factors in the
>model.
>I'd appreciate your input on this.
>Chang